Online Casino Paysafe - Craps Virtual Player
Transcription
Online Casino Paysafe - Craps Virtual Player
By William I. Orr, WGSAI 48 Campbell Lane, Menlo Park, California 94025 linear amplifier design The designer of a linear amplifier should be concerned with the proper potentials required to make the power tube operate in a linear manner. The word linear implies that the output signal of the amplifier is an amplified replica of the input signal. There's no such thing as a perfect linear amplifier, and the designer's problem is to make the practical amplifier ( i e . , the amplifier that can be built) as linear as possible. When a linear amplifier is driven by a complex signal, such as the human voice, nonlinearity results in intermodulation distortion. This unpleasant form of distortion creates a broad, raspy signal that throws annoying "buckshot" into adjacent channels. Proper design and operation of a linear amplifier reduces this distortion to a minimum. amplifier circuit and mode There's a lot of confusion with regard to the so) called "grounded-grid" amplifier. Rf power ampli- fiers are classified according to circuitry and mode of operation. The two classifications should not be confused with one another. For Amateur service, the two most popular circuits are the grid-driven circuit and the cathode-driven circuit. As shown in fig. 1, the circuits are remarkably similar, the most obvious difference being the placement of the ground point in relation to the input and output circuits. The mode of operation refers to the dynamic operating characteristics of the tube (class AB1, class B, or class C).Characteristics of the classes are given in reference material listed at the end of this article. For linear service, the power tube amplifier is commonly run in either class AB1 or class B service. Thus, modern equipment may have an intermix of circuitry and mode - the cathode-driven amplifier may be operated in a class AB1 mode, for example, or the griddriven amplifier may be operated in the class B mode. So far, I've not discussed the popular groundedgrid amplifier. This is a sloppy term which usually refers to a- cathode-driven amplifier, working in the class B mode. "Grounded grid" implies cathode drive, but in such a circuit the grid may not necessarily be at dc ground potential, especially with respect to screen voltage (see fig. 2). Rf ground and dc ground are not always the same in a linear amplifier, and most circuit engineers shudder at the use of the term. amplifier plate circuit While this series of articles concerns itself with linear, cathode-driven-amplifier design, the remarks about the plate circuit apply equally well t o grid-driven amplifiers. It is desirable to operate any linear amplifier with a very minimum of intermodulation distortion, with high-plate efficiency, and with high power gain. The latter is especially important, as it affords maximum power output with a given amount of drive power. The class B mode of operation meets these requirements. Shown in fig. 3 is a graphical representation of a class B amplifier, showing the operating cycle of the tube. This is the portion of the electrical cycle over which the tube grid is driven positive (approaching +e) with respect to the cathode (or the cathode driven negative with respect t o the grid). When the grid potential is highly negative with respect to the cathode (approaching -el, the tube is cut off and is inoperative. In the class B amplifier, the operating cycle is about one-half the electrical cycle, or approximately 180 degrees. The transfer curve plot shown indicates that the tube delivers power only over onehalf of the electrical cycle and is cut-off during the other half of the cycle. Does this mean that the out- . . ' put signal consists of half-sine waves as Shown, a"d'?: . - ' sf&. . is therefore highly distorted? Not at all. The amplifier plate circuit (often called the tank circuit) saves the day, since the energy storage ability ,. ' (Q) of the circuit balances the energy between the halves of the cycle, much as the flywheel stores energy during the operating cycles of a gasoline engine. The plate circuit must, therefore, be designed t o have sufficient 4, or energy storage, for good operation. A Q value of 12 is commonly used for linear amplifier service, as it provides ample energy storage and at the same time provides reasonable reduction of harmonics generated in the amplifier. - A By William I. Orr, WGSAI, 48 Campbell Lane, Menlo Park, California 94025 CATHODE DRIVEN CIRCUIT SUPPLY GRID DRIVEN CIRCUIT (neutrol~zingommtfledl SUPPLY SUPPLY SUPPLY SAFETY RESISTOR + M DC GROUND OC GROUND fig. 1. A comparison between grid-driven and cathode-driven amplifiers. Rf and dc circuits have been simplified for clarity. I n both cases, the grid- and plate-current meters are placed i n the ground return circuits t o remove any dangerous voltage f r o m the meter movement. This, however, places the plate supply above dc ground by virtue of the voltage across the plate meter. I f the meter coil should open, the negative lead of the supply rises t o the value of the plate voltage. As a safety factor, a wirewound resistor is usually placed across the plate meter, and often the grid meter. The circuit configuration determines the difference between cathode- and grid-driven service. The applied voltages determine the mode of operation. a A rigorous design of the plate circuit calls for manipulation of the plate voltage and current to determine the operating parameters of the tube. The results of these tedious calculations can be summed up in simple formulas that provide the designer with circuit data in everyday terms. A network is required that matches the plate load impedance of the power tube to the characteristic impedance of the transmission line, while at the same time maintaining a Qvalue of 12. The popular pi network can do the job. The plate load impedance (ZL) for a class B rf amplifier can be closely approximated by: load impedance (ohms) - up until the sloping line denoting a particular Amateur band is intersected. The value of the component is then read horizontally off the y axis. For example, the required inductance for a plate load of 1560 ohms for the 15 meter band is about one microhenry - as close as the graph can be read. Note that capacitor C1 is commonly referred to as the tuning capacitor and C2 the loading capacitor. The graph for C2 tells us that the pi network cannot cope with impedance transformation values much greater than 100-to-1 at this value of Q. Note how the curves bunch together and "fall-off the graph" at plate impedances much higher than 5000 ohms. plate uoltage 2 x peak dc plate current (amperes) As an example, a pi network is to be used to match a pair of 3-5002 tubes to a %-ohm transmission line. The tubes operate with 2500 volts plate potential with a peak dc plate current of 800 mA (0.8 amp) for a PEP input of 2 kW. load impedance = 2500 - 2 X 0.8 TETROOE TUBE LZ Thus, the pi network plate circuit has to match a load impedance of 1560 ohms to a %-ohm termination. designing the plate circuit network The approximate values of the pi network can be determined from three simple graphs, The plate inductance from fig. 4, the tuning capacitance (C1) from fig. 5, and the loading capacitance (c2) from fig. 6. The graphs are entered at the x axis and read + - SCREEN SUPPLY BUS SUPPLY + MI- M 1560ohms R F GROUND f c- LI 2 ;fi GROUND ' t - fig. 2. Diagram of the so-called "grounded-grid" amplifier. The grid and screen elements are bypassed t o ground as far as rf is concerned, but each element has normal operating voltages applied and are "above ground" as far as dc is concerned. Metering is inserted i n the supply return leads t o dc ground. Rf ground is placed at the positive screen voltage level. This eliminates the screen bypass capacitor, a tricky component that often causes circuit instability a t the higher frequencies. A more accurate, computer-derived summary of pi network values is given in table 1. Note that, for a given plate impedance, when the operating frequency is doubled the capacitance and inductance values are halved. (Fifteen- and forty-meter constants are related by a factor of three as 21 MHz is the third harmonic of 7 MHz.) coil winding Winding plate coil L1 to a given value of inductance takes an inductance meter, ora degree of experPLATE CURRENT AXIS a G TRANSFER CURVE PLATE SIGNAL fig. 4. Plot of the plate inductance vs. plate load impedance for the high frequency Amateur bands (Q = 12). POINT(E1 TUBE CUTOFF W R T I O N O F DRIVE CYCLE - 1 -- -- I I I / 2 CYCLE GRID VOLTAGE AXIS + e- GRID DRIVE VOLTAGE I * INSTANNNEOUS PLATE CURRENT E = BIASPOINT to= WIESCENT ON PLATE TRANSFER CURRENT CURVE I -. I I I I I DRIVE SIGNAL I 1 I Ce fig. 3. Transfer curve and operating cycle for a class B amplifier. The transfer curve is determined by a static test of the tube where plate current is plotted against grid bias. Once the transfer curve is established, the operating cycle may be determined. The sine wave drive signal (e) is drawn about the bias line, determining both the zero-signal plate current (i,) and the peak plate current (i,,,). Note that when the grid driving signal swings negative, no plate current is drawn and the tube is cut-off for one-half cycle. Pulses of plate current only appear when the drive signal is positive w i t h respect t o the bias voltage. Thus, the output waveform of a class B rf amplifier consists of a series of half-cycles, much i n the manner of a half-wave rectifier. The distorted waveform is restored t o a sine wave by the plate tank circuit which, by virtue of its Q, or flywheel effect, stores energy on the active half of the cycle and releases it on the inactive half. Circuit engineers, working from a transfer curve, can determine actual dc operating potentials f o r a linear amplifier. ple slide rule providing direct read-out of the coil dimensions if the inductance is known. It takes the hard work out of designing coils. Once the plate circuit has been designed and built, it is a good idea to "breadboard" it up and check it out with a dip-meter before the connections are finally soldered. Coil taps may have to be moved a bit to compensate for capacitance of the components to the chassis and adjacent parts. amplifier-cathode circuit The cathode-input circuit provides an impedance match between the 50-ohm coaxial output circuit of the driverlexciter and the input impedance of the cathode-driven amplifier (see table 2). The input im- tise and a dip-meter. A simple formula for calculating inductance when the coil dimensions are known is: Inductance (fl)= R2N2 gR + IOS where R is the radius of the coil in inches S is the length of the coil winding in inches N is the number of turns These calculations have been simplified in the ARRL type-A "Lightning Calculator," which is a sim- 36 july 1979 fig. 5. plot of the tuning capacitance (C1) vs. plate load impedance ( Q = 12). pedance (Zt) of a cathode-driven tube is related to the ratio of the peak cathode signal voltage to the peak cathode current (sum of grid and plate currents), and is commonly given in the tube data sheet. For the 3-5002 at 2500 volts, it is about 110 ohms. And for two tubes in parallel, it is about 55 ohms, but only over the operating cycle. It is tempting to jump to the conclusion that if the amplifier input impedance is about 55 ohms and the coaxial line impedance driving it is 50 ohms, that no cathode impedance matching circuit is required. In fact, many commercially manufactured amplifiers leave it out for economy's sake. This omission is poor engineering practice, as the circuit Q is required in the cathode circuit as well as in the plate circuit. Omission of the cathode-tuned circuit can lead to distortion of the driving signal, increased intermodulation distortion, reduced amplifier efficiency, and driver loading problems. A circuit Qof 2 is adequate, and a simple rule of thumb is that the network circuit capacitances at resonance should be about 20 pF per meter of wavelength for one-to-one impedance transformation. practical amplifier circuit Armed with the information discussed so far, it is possible t o draw up a schematic for a cathode driven, 2-kW PEP linear amplifier using two 3-5002 tubes in parallel (see fig. 7 ) . This is a true "grounded-grid" circuit, as the grids are at both dc and rf ground potential. fig. 6. Plot of the loading capacitance (C2) vs. plate load impedance (Q = 121. Note that plate and grid currents are measured in the cathode return circuit. This requires the amplifier plate power supply to "float" a little above ground potential in order to insert a meter in the negative lead to measure plate current. This removes the lethal plate voltage from the meter. The grid meter is out of the critical rf ground return path, which simplifies the metering circuit. A filament voltmeter is included. Filament voltage should be held to within LOAD = 5 0 OHM table 1. Computer-derived values for a pi network having a Q o f 12 and working into a 50ohm load. Values for C1 include the output capacitance of the tubes. These values are taken from a computer program derived by Bob Sutherland, W6PO. ZL plate load impedance (ohms) component band 160 80 C1 40 20 15 10 1000 1060 546 273 136 91 68 1500 690 364 182 91 61 45 2000 531 273 136 68 45 34 2500 430 220 110 55 37 30 3000 354 182 91 45 30 23 3500 309 159 80 40 26 20 4000 265 136 68 34 23 17 5000 212 109 55 27 18 14 f 5 per cent of 5 volts, and it is prudent to monitor this voltage when expensive tubes are used. A plate voltmeter may be included in the amplifier, but it is easier t o place it in the power supply. Amplifier standby plate current is reduced by means of a 10-kilohm, 25-watt cathode resistor which is shorted out by the VOX relay of the exciter, causing the tubes to operate at the proper resting plate current when the amplifier is on the air. A zener diode is placed in series with the cathode dc return path t o reduce the quiescent plate current during amplifier operation. A %-ohm wirewound resistor from the negative side of the plate supply to ground makes certain that the negative supply terminal does not rise to the value of the plate voltage if the positive side of the supply is accidentally shorted to ground. Two reverse-connected diodes are shunted across the safety resistor to limit any transient surges under a shorted condition which might cause wiring insula- tion breakdown. In addition, the diodes protect the meters from transient currents. A resistor across the zener diode provides a constant load for it and prevents cathode voltage from soaring if the zener safety fuse opens. Note that a 10-ohm, %-watt wirewound resistor is placed in series with the B-plus lead to the plate rf choke. This resistor serves as a vhf choke to suppress harmonic currents in the power lead and also protects the tube and associated circuitry in case of a flash-over in the tube or plate circuit. The trernendous amount of energy stored in the power supply is instantaneously "dumped" into the amplifier when a FOR 2," - 5 0 OHM 0.2 table 2. The pi-network circuit for a cathode-driven amplifier. This chart provides approximate values for the components of the cathode circuit. Capacitors should be I - k V silver mica or equivalent. The inductor can be wound on a slug-tuned form. Value of C2 should take into account the cathode-grid capacitance of the tube which appears in parallel with C2 (information is from a computer program by W6POl. cathode z,(m 20 38 july 1979 band Cl(pFI C2(pF) L(pH1 160 80 40 20 15 10 3300 1700 900 440 300 220 4100 2120 1120 560 370 275 2.50 1.34 0.68 0.33 0.22 0.16 cathode Z,(Q) 75 band Cl(pF) C2(pF) 160 80 40 20 15 10 3300 1700 900 440 300 220 2870 1540 770 380 250 180 L~(PHI 3.81 2.05 1.03 0.51 0.34 0.25 C6 0 1 SKY I / - - LP L3 - J2 OUTPUT RFC 3 'OM 40M 15M 0 01 001 001 001 d DANGER- HIGH VOLTAGE L - _ - 1 - - - ---------- 2 3 4 5 6 7 J 8 fig. 7 . Schematic diagram of t h e 3-5002 linear amplifier. C3 C4 C5 C6 250 pF, 4.5 kV plate spacing - Johnson 154-16 500 pF, 4.5 kV 1000 pF, 500 volt plate spacing 0.001 ,F, 5 kV - Centralab 858s-100 C7, C8 500 pF, 10 kV TV-type "door knob" C9-C14 0.01 ,F, 500 volt mica capacitor. Ceramic disc is a suitable substitute if rated 1 kV. PC 1 Three 100-ohm, 2-watt resistors in parallel PC 2 Three turns of no. 14 AWG (1.6 m m ) wound with 12.5m m (0.5-inch) diameter and 19-mm (0.75-inch) length connected in parallel with the resistors. The coil may be wound around one of the resistors. flash-over occurs, and much of this destructive energy is dissipated in the resistor. Many modern-generation Amateurs have never worked with equipment operating at voltages higher than 12 volts. This amplifier, with the high-voltage plate supply, is positively lethal and the operator can be killed if his hands are inside the unit when the high voltage is on. It is imperative, therefore, that safety switches be incorporated in the amplifier design. It is poor engineering practice to leave these devices out! S4 isa normally open, pushbuttondevice that isclosed only when the lid is placed on the amplifier enclosure. S3 is a shorting switch that shorts the high voltage to ground when the lid is removed. Construction of this special switch will be covered in a future article. Always remember - high voltage kills! Take necessary precautions. RFC 1 RFC 2 RFC 3 TI Blower 50 pH; 14 bifilar turns of no. 10 AWG (2.6 rnm) enameled wire wound on ferrite core 12.5 cm (5 inches) long and 12.5 c m (0.5 inch) in diameter (Indiana General CF-503 or equivalent). 100 pH, 1 ampere dc; 112 turns no. 26 AWG (0.4 mm) spacewound wire diameter on 2.5 c m (1 inch) ceramic f o r m 15 c m ( 6 inches) long (Centralab X-3022H insulator). Series resonant at 24.5 MHz with terminals shorted ( B 8 W 800). 2.5 mH, 100 mA 5 volts at 30 amps (Chicago-Standard P-4648) 13 cu. ft./min. Use a no. 3 impeller at 3100 rpm (Ripley 8472, Dayton 1C-180, or Redmond AK-2H-OlAX) Although not shown on the schematic, it is a good idea to use a filament transformer having a primary winding tapped for 105, 115, and 125 volts. This provides a plus or minus ten per cent adjustment from a normal line voltage of 115 volts. If a closer filament adjustment is desirable, the transformer can be run on the 105 volt tap with a rheostat in series with the primary winding to place the filament voltage "on the nose." The plus and minus leads to the high voltage supply should be run through high-voltage connectors and high-voltage cable. Test prod wire having a 10kV breakdown is satisfactory. As an alternative, RG58lU coaxial cable can be used for high-voltage leads along with PL-259 plugs and reducers and SO-239 receptacles. The shield of the coaxial line is grounded by the connectors. ham radio